This class provides an interface to enumerating an IP range
This class uses start,stop pairs to represent ranges of addresses. This is very efficient for large numbers of consecutive addresses, and not show-stoppingly inefficient when storing a bunch of non-consecutive addresses, which should be a somewhat unusual case.
Calls the given block with each address. This is basically a wrapper for next_ip
# File lib/rex/socket/range_walker.rb, line 216 def each(&block) while (ip = next_ip) block.call(ip) end end
Returns an array with one element, a Range defined by the given CIDR block.
# File lib/rex/socket/range_walker.rb, line 226 def expand_cidr(arg) start,stop = Rex::Socket.cidr_crack(arg) if !start or !stop return false end range = Range.new range.start = Rex::Socket.addr_atoi(start) range.stop = Rex::Socket.addr_atoi(stop) range.ipv6 = (arg.include?(":")) range.options = {} return range end
Expands an nmap-style host range x.x.x.x where x can be simply "*" which means 0-255 or any combination and repitition of:
i,n n-m i,n-m n-m,i
ensuring that n is never greater than m.
non-unique elements will be removed
e.g.: 10.1.1.1-3,2-2,2 => ["10.1.1.1", "10.1.1.2", "10.1.1.3"] 10.1.1.1-3,7 => ["10.1.1.1", "10.1.1.2", "10.1.1.3", "10.1.1.7"]
Returns an array of Ranges
# File lib/rex/socket/range_walker.rb, line 256 def expand_nmap(arg) # Can't really do anything with IPv6 return false if arg.include?(":") # nmap calls these errors, but it's hard to catch them with our # splitting below, so short-cut them here return false if arg.include?(",-") or arg.include?("-,") bytes = [] sections = arg.split('.') if sections.length != 4 # Too many or not enough dots return false end sections.each { |section| if section.empty? # pretty sure this is an unintentional artifact of the C # functions that turn strings into ints, but it sort of makes # sense, so why not # "10...1" => "10.0.0.1" section = "0" end if section == "*" # I think this ought to be 1-254, but this is how nmap does it. section = "0-255" elsif section.include?("*") return false end # Break down the sections into ranges like so # "1-3,5-7" => ["1-3", "5-7"] ranges = section.split(',', -1) sets = [] ranges.each { |r| bounds = [] if r.include?('-') # Then it's an actual range, break it down into start,stop # pairs: # "1-3" => [ 1, 3 ] # if the lower bound is empty, start at 0 # if the upper bound is empty, stop at 255 # bounds = r.split('-', -1) return false if (bounds.length > 2) bounds[0] = 0 if bounds[0].nil? or bounds[0].empty? bounds[1] = 255 if bounds[1].nil? or bounds[1].empty? bounds.map!{|b| b.to_i} return false if bounds[0] > bounds[1] else # Then it's a single value bounds[0] = r.to_i end return false if bounds[0] > 255 or (bounds[1] and bounds[1] > 255) return false if bounds[1] and bounds[0] > bounds[1] if bounds[1] bounds[0].upto(bounds[1]) do |i| sets.push(i) end elsif bounds[0] sets.push(bounds[0]) end } bytes.push(sets.sort.uniq) } # # Combinitorically squish all of the quads together into a big list of # ip addresses, stored as ints # # e.g.: # [[1],[1],[1,2],[1,2]] # => # [atoi("1.1.1.1"),atoi("1.1.1.2"),atoi("1.1.2.1"),atoi("1.1.2.2")] addrs = [] for a in bytes[0] for b in bytes[1] for c in bytes[2] for d in bytes[3] ip = (a << 24) + (b << 16) + (c << 8) + d addrs.push ip end end end end addrs.sort! addrs.uniq! rng = Range.new rng.ipv6 = false rng.options = {} rng.start = addrs[0] ranges = [] 1.upto(addrs.length - 1) do |idx| if addrs[idx - 1] + 1 == addrs[idx] # Then this address is contained in the current range next else # Then this address is the upper bound for the current range rng.stop = addrs[idx - 1] ranges.push(rng.dup) rng.start = addrs[idx] end end rng.stop = addrs[addrs.length - 1] ranges.push(rng.dup) return ranges end
Returns true if the argument is an ip address that falls within any of the stored ranges.
# File lib/rex/socket/range_walker.rb, line 183 def include?(addr) return false if not @ranges if (addr.is_a? String) addr = Rex::Socket.addr_atoi(addr) end @ranges.map { |r| if r[0] <= addr and addr <= r[1] return true end } return false end
Returns true if this RangeWalker includes all of the addresses in the given RangeWalker
# File lib/rex/socket/range_walker.rb, line 200 def include_range?(range_walker) return false if ((not @ranges) or @ranges.empty?) return false if not range_walker.ranges range_walker.ranges.all? do |start, stop| ranges.any? do |self_start, self_stop| r = (self_start..self_stop) r.include?(start) and r.include?(stop) end end end
Returns the next IP address.
# File lib/rex/socket/range_walker.rb, line 156 def next_ip return false if not valid? if (@curr_addr > @ranges[@curr_range][1]) if (@curr_range >= @ranges.length - 1) return nil end @curr_range += 1 @curr_addr = @ranges[@curr_range][0] end addr = Rex::Socket.addr_itoa(@curr_addr, @ranges[@curr_range][2]) if @ranges[@curr_range][3][:scope_id] addr = addr + '%' + @ranges[@curr_range][3][:scope_id] end @curr_addr += 1 return addr end
Turn a human-readable range string into ranges we can step through one address at a time.
Allow the following formats:
"a.b.c.d e.f.g.h" "a.b.c.d, e.f.g.h"
where each chunk is CIDR notation, (e.g. '10.1.1.0/24') or a range in nmap format (see expand_nmap)
OR this format
"a.b.c.d-e.f.g.h"
where a.b.c.d and e.f.g.h are single IPs and the second must be bigger than the first.
# File lib/rex/socket/range_walker.rb, line 52 def parse(parseme) return nil if not parseme ranges = [] parseme.split(', ').map{ |a| a.split(' ') }.flatten.each { |arg| opts = {} # Handle IPv6 first (support ranges, but not CIDR) if arg.include?(":") addrs = arg.split('-', 2) # Handle a single address if addrs.length == 1 addr, scope_id = addrs[0].split('%') opts[:scope_id] = scope_id if scope_id return false unless Rex::Socket.is_ipv6?(addr) addr = Rex::Socket.addr_atoi(addr) ranges.push [addr, addr, true, opts] next end addr1, scope_id = addrs[0].split('%') opts[:scope_id] = scope_id if scope_id addr2, scope_id = addrs[0].split('%') ( opts[:scope_id] ||= scope_id ) if scope_id return false if not (Rex::Socket.is_ipv6?(addr1) and Rex::Socket.is_ipv6?(addr2)) # Handle IPv6 ranges in the form of 2001::1-2001::10 addr1 = Rex::Socket.addr_atoi(addr1) addr2 = Rex::Socket.addr_atoi(addr2) ranges.push [addr1, addr2, true, opts] next # Handle IPv4 CIDR elsif arg.include?("/") # Then it's CIDR notation and needs special case return false if arg =~ /[,-]/ # Improper CIDR notation (can't mix with 1,3 or 1-3 style IP ranges) return false if arg.scan("/").size > 1 # ..but there are too many slashes ip_part,mask_part = arg.split("/") return false if ip_part.nil? or ip_part.empty? or mask_part.nil? or mask_part.empty? return false if mask_part !~ /^[0-9]{1,2}$/ # Illegal mask -- numerals only return false if mask_part.to_i > 32 # This too -- between 0 and 32. begin Rex::Socket.addr_atoi(ip_part) # This allows for "www.metasploit.com/24" which is fun. rescue Resolv::ResolvError return false # Can't resolve the ip_part, so bail. end expanded = expand_cidr(arg) if expanded ranges.push(expanded) else return false end # Handle hostnames elsif arg =~ /[^-0-9,.*]/ # Then it's a domain name and we should send it on to addr_atoi # unmolested to force a DNS lookup. Rex::Socket.addr_atoi_list(arg).each { |addr| ranges.push [addr, addr, false, opts] } # Handle IPv4 ranges elsif arg =~ /^([0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3})-([0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3})$/ # Then it's in the format of 1.2.3.4-5.6.7.8 # Note, this will /not/ deal with DNS names, or the fancy/obscure 10...1-10...2 begin addrs = [Rex::Socket.addr_atoi($1), Rex::Socket.addr_atoi($2)] return false if addrs[0] > addrs[1] # The end is greater than the beginning. ranges.push [addrs[0], addrs[1], false, opts] rescue Resolv::ResolvError # Something's broken, forget it. return false end else # Returns an array of ranges expanded = expand_nmap(arg) if expanded expanded.each { |r| ranges.push(r) } end end } # Remove any duplicate ranges ranges = ranges.uniq return ranges end
Generated with the Darkfish Rdoc Generator 2.